Linux 6.13 AMD EPYC Performance

Benchmarks for a future article. AMD EPYC 9575F 64-Core testing with a Supermicro Super Server H13SSL-N v1.01 (3.0 BIOS) and ASPEED on Ubuntu 24.10 via the Phoronix Test Suite.

Compare your own system(s) to this result file with the Phoronix Test Suite by running the command: phoronix-test-suite benchmark 2412073-NE-LINUX613A93
Jump To Table - Results

View

Do Not Show Noisy Results
Do Not Show Results With Incomplete Data
Do Not Show Results With Little Change/Spread
List Notable Results
Show Result Confidence Charts
Allow Limiting Results To Certain Suite(s)

Statistics

Show Overall Harmonic Mean(s)
Show Overall Geometric Mean
Show Wins / Losses Counts (Pie Chart)
Normalize Results
Remove Outliers Before Calculating Averages

Graph Settings

Force Line Graphs Where Applicable
Convert To Scalar Where Applicable
Prefer Vertical Bar Graphs

Multi-Way Comparison

Condense Multi-Option Tests Into Single Result Graphs

Table

Show Detailed System Result Table

Run Management

Highlight
Result
Toggle/Hide
Result
Result
Identifier
Performance Per
Dollar
Date
Run
  Test
  Duration
v6.11
December 07
  7 Hours, 42 Minutes
v6.12
December 06
  7 Hours, 21 Minutes
v6.13 7 Dec
December 07
  7 Hours, 41 Minutes
Invert Behavior (Only Show Selected Data)
  7 Hours, 35 Minutes

Only show results where is faster than
Only show results matching title/arguments (delimit multiple options with a comma):
Do not show results matching title/arguments (delimit multiple options with a comma):


Linux 6.13 AMD EPYC PerformanceProcessorMotherboardChipsetMemoryDiskGraphicsNetworkOSKernelDesktopDisplay ServerCompilerFile-SystemScreen Resolutionv6.11v6.12v6.13 7 DecAMD EPYC 9575F 64-Core @ 3.30GHz (64 Cores / 128 Threads)Supermicro Super Server H13SSL-N v1.01 (3.0 BIOS)AMD 1Ah12 x 64GB DDR5-6000MT/s Micron MTC40F2046S1RC64BDY QSFF3201GB Micron_7450_MTFDKCB3T2TFSASPEED2 x Broadcom NetXtreme BCM5720 PCIeUbuntu 24.106.11.0-phx (x86_64)GNOME Shell 47.0X ServerGCC 14.2.0ext41920x12006.12.0-phx (x86_64)1024x7686.13.0-rc1-phx (x86_64)OpenBenchmarking.orgKernel Details- Transparent Huge Pages: madviseCompiler Details- --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2,rust --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-14-zdkDXv/gcc-14-14.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-14-zdkDXv/gcc-14-14.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v Processor Details- Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xb002116Java Details- OpenJDK Runtime Environment (build 21.0.5+11-Ubuntu-1ubuntu124.10)Python Details- Python 3.12.7Security Details- gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS; IBPB: conditional; STIBP: always-on; RSB filling; PBRSB-eIBRS: Not affected; BHI: Not affected + srbds: Not affected + tsx_async_abort: Not affected

v6.11v6.12v6.13 7 DecResult OverviewPhoronix Test Suite100%110%121%131%141%PalabosStress-NGMariaDBApache CouchDBRELIONPostgreSQLMemcachedApache CassandraSVT-AV1LaghosBlenderLlamafileClickHousePrimesieveOpenSSLOpenVINO GenAITimed LLVM CompilationTimed Linux Kernel CompilationDaCapo Benchmark

Linux 6.13 AMD EPYC Performancecouchdb: 300 - 3000 - 30couchdb: 500 - 1000 - 30relion: Basic - CPUpalabos: 1000mariadb: oltp_update_index - 128mariadb: oltp_read_only - 128mariadb: oltp_read_write - 128mariadb: oltp_write_only - 128build-linux-kernel: allmodconfigopenssl: ChaCha20openssl: ChaCha20-Poly1305openssl: AES-256-GCMopenssl: AES-128-GCMbuild-llvm: Unix Makefilesblender: Barbershop - CPU-Onlycouchdb: 300 - 1000 - 30svt-av1: Preset 3 - Bosphorus 4Kpgbench: 100 - 1000 - Read Write - Average Latencypgbench: 100 - 1000 - Read Writepgbench: 100 - 800 - Read Write - Average Latencypgbench: 100 - 800 - Read Writepgbench: 100 - 800 - Read Only - Average Latencypgbench: 100 - 800 - Read Onlypgbench: 100 - 1000 - Read Only - Average Latencypgbench: 100 - 1000 - Read Onlypalabos: 500cassandra: Writesclickhouse: 100M Rows Hits Dataset, Third Runclickhouse: 100M Rows Hits Dataset, Second Runclickhouse: 100M Rows Hits Dataset, First Run / Cold Cachebuild-llvm: Ninjadacapobench: Tradebeansllamafile: wizardcoder-python-34b-v1.0.Q6_K - Prompt Processing 2048stress-ng: Semaphoresmemcached: 1:100memcached: 1:10memcached: 1:5stress-ng: Context Switchinglaghos: Sedov Blast Wave, ube_922_hex.meshllamafile: mistral-7b-instruct-v0.2.Q5_K_M - Prompt Processing 2048blender: Pabellon Barcelona - CPU-Onlyblender: Classroom - CPU-Onlysvt-av1: Preset 5 - Bosphorus 4Kllamafile: wizardcoder-python-34b-v1.0.Q6_K - Text Generation 128llamafile: wizardcoder-python-34b-v1.0.Q6_K - Prompt Processing 1024dacapobench: Tradesoaplaghos: Triple Point Problemdacapobench: Apache Lucene Search Engineopenvino-genai: Falcon-7b-instruct-int4-ov - CPU - Time Per Output Tokenopenvino-genai: Falcon-7b-instruct-int4-ov - CPU - Time To First Tokenopenvino-genai: Falcon-7b-instruct-int4-ov - CPUstress-ng: NUMAstress-ng: Mixed Schedulerstress-ng: Futexstress-ng: MEMFDstress-ng: Socket Activitystress-ng: Mutexllamafile: mistral-7b-instruct-v0.2.Q5_K_M - Text Generation 128llamafile: TinyLlama-1.1B-Chat-v1.0.BF16 - Text Generation 128build-linux-kernel: defconfigopenvino-genai: Gemma-7b-int4-ov - CPU - Time Per Output Tokenopenvino-genai: Gemma-7b-int4-ov - CPU - Time To First Tokenopenvino-genai: Gemma-7b-int4-ov - CPUdacapobench: Avrora AVR Simulation Frameworkprimesieve: 1e13llamafile: mistral-7b-instruct-v0.2.Q5_K_M - Prompt Processing 1024llamafile: wizardcoder-python-34b-v1.0.Q6_K - Text Generation 16dacapobench: Eclipsellamafile: Llama-3.2-3B-Instruct.Q6_K - Prompt Processing 2048dacapobench: H2 Database Engineblender: Fishy Cat - CPU-Onlydacapobench: jMonkeyEngineblender: Junkshop - CPU-Onlyllamafile: wizardcoder-python-34b-v1.0.Q6_K - Prompt Processing 512dacapobench: PMD Source Code Analyzerllamafile: Llama-3.2-3B-Instruct.Q6_K - Text Generation 128dacapobench: Apache Kafkadacapobench: Apache Lucene Search Indexblender: BMW27 - CPU-Onlyllamafile: mistral-7b-instruct-v0.2.Q5_K_M - Text Generation 16openvino-genai: TinyLlama-1.1B-Chat-v1.0 - CPU - Time Per Output Tokenopenvino-genai: TinyLlama-1.1B-Chat-v1.0 - CPU - Time To First Tokenopenvino-genai: TinyLlama-1.1B-Chat-v1.0 - CPUdacapobench: Jythonsvt-av1: Preset 8 - Bosphorus 4Kllamafile: TinyLlama-1.1B-Chat-v1.0.BF16 - Prompt Processing 2048dacapobench: BioJava Biological Data Frameworkllamafile: mistral-7b-instruct-v0.2.Q5_K_M - Prompt Processing 512dacapobench: Spring Bootopenvino-genai: Phi-3-mini-128k-instruct-int4-ov - CPU - Time Per Output Tokenopenvino-genai: Phi-3-mini-128k-instruct-int4-ov - CPU - Time To First Tokenopenvino-genai: Phi-3-mini-128k-instruct-int4-ov - CPUllamafile: Llama-3.2-3B-Instruct.Q6_K - Prompt Processing 1024llamafile: Llama-3.2-3B-Instruct.Q6_K - Text Generation 16dacapobench: GraphChillamafile: wizardcoder-python-34b-v1.0.Q6_K - Prompt Processing 256llamafile: TinyLlama-1.1B-Chat-v1.0.BF16 - Text Generation 16llamafile: mistral-7b-instruct-v0.2.Q5_K_M - Prompt Processing 256svt-av1: Preset 13 - Bosphorus 4Kllamafile: TinyLlama-1.1B-Chat-v1.0.BF16 - Prompt Processing 1024dacapobench: Apache Tomcatllamafile: Llama-3.2-3B-Instruct.Q6_K - Prompt Processing 512dacapobench: Zxing 1D/2D Barcode Image Processingdacapobench: Batik SVG Toolkitdacapobench: Apache Xalan XSLTllamafile: Llama-3.2-3B-Instruct.Q6_K - Prompt Processing 256llamafile: TinyLlama-1.1B-Chat-v1.0.BF16 - Prompt Processing 512dacapobench: FOP Print Formatterprimesieve: 1e12llamafile: TinyLlama-1.1B-Chat-v1.0.BF16 - Prompt Processing 256v6.11v6.12v6.13 7 Dec363.527354.526176.673481.48919209843688183693412967194.338718106876590487834985450333162995897359931303710159.819149.30103.22816.8808.4981176776.3051268770.16747836480.2174621379529.232483952801.19804.06770.47101.821499912288233159294.1412696519.857421710.244115352.4936748330.54561.913276847.9042.0360.17115.9061443077297.99412017.6330.3956.732071.9879773.374154651.852054.7047424.8745376718.9468.13159.9021.79220.3630.3949.13235425.0361638414.98634932768210421.66679120.2430721085112.565070229415.1964.8712.7615.8678.363782199.2093276840588192272415.1822.6665.8916384108.2220931536152.074096466.6141638410538192522923743409681923391.9994096371.288333.178179.868655.18418353743581181192399417193.202721616676910490511085203334486696117361564053870160.302148.82104.16316.9908.9271120256.5501221810.16648156250.2114732753775.110491186799.99801.90765.59101.599496812288251549880.0113294800.597297642.214064048.1455755486.32565.103276847.7041.7460.38115.8261443035303.11432617.8030.0056.212298.6582674.984313954.942010.9947501.0445398166.1467.71159.6021.84820.3730.2749.09230324.9721638415.01628032768204921.49679720.2130721167111.985045225815.0864.8012.7915.8978.183755200.3163276840718192225915.1922.7365.8216384107.9520831536152.894096468.6661638410388192551937741409681923351.9904096360.826311.038172.482509.12615830442936179909386983192.923719980763180489486102253334428157450361324888880160.559148.16100.14017.0089.0821101126.6281207030.16548292550.2114735578764.024487743807.27803.49774.01101.725495712288246944948.4613514659.677400996.984099640.3153038526.00565.143276847.3941.4960.82115.7761443049300.72363117.7730.1856.282385.5981582.584398887.772125.1347238.7043186401.8465.45156.8221.85520.5430.4848.70228324.9391638414.90633132768201021.35679320.2030721089109.605057227815.0264.1512.8816.0577.673763200.8233276840198192265115.1722.3865.9116384102.2121021536147.984096480.9121638410558192493938742409681923381.9884096OpenBenchmarking.org

Apache CouchDB

OpenBenchmarking.orgSeconds, Fewer Is BetterApache CouchDB 3.4.1Bulk Size: 300 - Inserts: 3000 - Rounds: 30v6.12v6.11v6.13 7 Dec80160240320400SE +/- 12.48, N = 9SE +/- 6.26, N = 9SE +/- 3.45, N = 9371.29363.53360.831. (CXX) g++ options: -flto -lstdc++ -shared -lei

OpenBenchmarking.orgSeconds, Fewer Is BetterApache CouchDB 3.4.1Bulk Size: 500 - Inserts: 1000 - Rounds: 30v6.11v6.12v6.13 7 Dec80160240320400SE +/- 62.71, N = 8SE +/- 36.20, N = 9SE +/- 29.09, N = 7354.53333.18311.041. (CXX) g++ options: -flto -lstdc++ -shared -lei

RELION

OpenBenchmarking.orgSeconds, Fewer Is BetterRELION 5.0Test: Basic - Device: CPUv6.12v6.11v6.13 7 Dec4080120160200SE +/- 0.79, N = 3SE +/- 1.95, N = 3SE +/- 3.31, N = 12179.87176.67172.481. (CXX) g++ options: -fPIC -std=c++14 -fopenmp -O3 -rdynamic -ldl -ltiff -lfftw3f -lfftw3 -lpng -ljpeg -lmpi_cxx -lmpi

Palabos

OpenBenchmarking.orgMega Site Updates Per Second, More Is BetterPalabos 2.3Grid Size: 1000v6.11v6.13 7 Decv6.12140280420560700SE +/- 3.07, N = 15SE +/- 21.98, N = 15SE +/- 36.69, N = 12481.49509.13655.181. (CXX) g++ options: -std=c++17 -pedantic -O3 -rdynamic -lcrypto -lcurl -lsz -lz -ldl -lm

MariaDB

This is a MariaDB MySQL database server benchmark making use of sysbench rather than the existing pts/mysqlslap test profile that uses MariaDB with mysqlslap/mariadb-slap as the benchmark driver. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB 11.5Test: oltp_update_index - Threads: 128v6.13 7 Decv6.12v6.1140K80K120K160K200KSE +/- 112.30, N = 3SE +/- 76.99, N = 3SE +/- 123.94, N = 31583041835371920981. (CXX) g++ options: -pie -fPIC -fstack-protector -O3 -lnuma -lpcre2-8 -lcrypt -laio -lz -lm -lssl -lcrypto -lpthread -ldl

OpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB 11.5Test: oltp_read_only - Threads: 128v6.13 7 Decv6.12v6.119K18K27K36K45KSE +/- 18.45, N = 3SE +/- 15.19, N = 3SE +/- 14.78, N = 34293643581436881. (CXX) g++ options: -pie -fPIC -fstack-protector -O3 -lnuma -lpcre2-8 -lcrypt -laio -lz -lm -lssl -lcrypto -lpthread -ldl

OpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB 11.5Test: oltp_read_write - Threads: 128v6.13 7 Decv6.12v6.1140K80K120K160K200KSE +/- 77.99, N = 3SE +/- 60.34, N = 3SE +/- 155.52, N = 31799091811921836931. (CXX) g++ options: -pie -fPIC -fstack-protector -O3 -lnuma -lpcre2-8 -lcrypt -laio -lz -lm -lssl -lcrypto -lpthread -ldl

OpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB 11.5Test: oltp_write_only - Threads: 128v6.13 7 Decv6.12v6.1190K180K270K360K450KSE +/- 154.44, N = 3SE +/- 474.96, N = 3SE +/- 286.92, N = 33869833994174129671. (CXX) g++ options: -pie -fPIC -fstack-protector -O3 -lnuma -lpcre2-8 -lcrypt -laio -lz -lm -lssl -lcrypto -lpthread -ldl

Timed Linux Kernel Compilation

This test times how long it takes to build the Linux kernel in a default configuration (defconfig) for the architecture being tested or alternatively an allmodconfig for building all possible kernel modules for the build. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Linux Kernel Compilation 6.8Build: allmodconfigv6.11v6.12v6.13 7 Dec4080120160200SE +/- 0.21, N = 3SE +/- 0.34, N = 3SE +/- 0.25, N = 3194.34193.20192.92

OpenSSL

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSLAlgorithm: ChaCha20v6.11v6.13 7 Decv6.12160000M320000M480000M640000M800000MSE +/- 652717270.91, N = 3SE +/- 626958911.68, N = 3SE +/- 1474303291.59, N = 37181068765907199807631807216166769101. OpenSSL 3.3.1 4 Jun 2024 (Library: OpenSSL 3.3.1 4 Jun 2024) - Additional Parameters: -engine qatengine -async_jobs 8

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSLAlgorithm: ChaCha20-Poly1305v6.11v6.13 7 Decv6.12110000M220000M330000M440000M550000MSE +/- 271433364.08, N = 3SE +/- 158903312.00, N = 3SE +/- 45557628.93, N = 34878349854504894861022534899042209371. OpenSSL 3.3.1 4 Jun 2024 (Library: OpenSSL 3.3.1 4 Jun 2024) - Additional Parameters: -engine qatengine -async_jobs 8

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSLAlgorithm: AES-256-GCMv6.11v6.12v6.13 7 Dec70000M140000M210000M280000M350000MSE +/- 105448439.59, N = 3SE +/- 189485712.03, N = 3SE +/- 134252719.45, N = 33331629958973342797221403344281574501. OpenSSL 3.3.1 4 Jun 2024 (Library: OpenSSL 3.3.1 4 Jun 2024) - Additional Parameters: -engine qatengine -async_jobs 8

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSLAlgorithm: AES-128-GCMv6.11v6.13 7 Decv6.1280000M160000M240000M320000M400000MSE +/- 130266226.11, N = 3SE +/- 88875000.56, N = 3SE +/- 210850977.29, N = 33599313037103613248888803615136146301. OpenSSL 3.3.1 4 Jun 2024 (Library: OpenSSL 3.3.1 4 Jun 2024) - Additional Parameters: -engine qatengine -async_jobs 8

Timed LLVM Compilation

This test times how long it takes to compile/build the LLVM compiler stack. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed LLVM Compilation 16.0Build System: Unix Makefilesv6.13 7 Decv6.12v6.114080120160200SE +/- 0.86, N = 3SE +/- 0.33, N = 3SE +/- 0.34, N = 3160.56160.30159.82

Blender

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.3Blend File: Barbershop - Compute: CPU-Onlyv6.11v6.12v6.13 7 Dec306090120150SE +/- 0.08, N = 3SE +/- 0.18, N = 3SE +/- 0.23, N = 3149.30148.82148.16

Apache CouchDB

OpenBenchmarking.orgSeconds, Fewer Is BetterApache CouchDB 3.4.1Bulk Size: 300 - Inserts: 1000 - Rounds: 30v6.12v6.11v6.13 7 Dec20406080100SE +/- 0.71, N = 3SE +/- 1.11, N = 3SE +/- 0.87, N = 7104.16103.23100.141. (CXX) g++ options: -flto -lstdc++ -shared -lei

SVT-AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.3Encoder Mode: Preset 3 - Input: Bosphorus 4Kv6.11v6.12v6.13 7 Dec48121620SE +/- 0.08, N = 3SE +/- 0.04, N = 3SE +/- 0.02, N = 316.8816.9917.011. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

PostgreSQL

OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL 17Scaling Factor: 100 - Clients: 1000 - Mode: Read Write - Average Latencyv6.13 7 Decv6.12v6.113691215SE +/- 0.042, N = 3SE +/- 0.048, N = 3SE +/- 0.020, N = 39.0828.9278.4981. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpq -lpgcommon -lpgport -lm

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL 17Scaling Factor: 100 - Clients: 1000 - Mode: Read Writev6.13 7 Decv6.12v6.1130K60K90K120K150KSE +/- 515.38, N = 3SE +/- 604.26, N = 3SE +/- 274.82, N = 31101121120251176771. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpq -lpgcommon -lpgport -lm

OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL 17Scaling Factor: 100 - Clients: 800 - Mode: Read Write - Average Latencyv6.13 7 Decv6.12v6.11246810SE +/- 0.030, N = 3SE +/- 0.082, N = 3SE +/- 0.030, N = 36.6286.5506.3051. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpq -lpgcommon -lpgport -lm

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL 17Scaling Factor: 100 - Clients: 800 - Mode: Read Writev6.13 7 Decv6.12v6.1130K60K90K120K150KSE +/- 536.64, N = 3SE +/- 1549.82, N = 3SE +/- 594.42, N = 31207031221811268771. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpq -lpgcommon -lpgport -lm

OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL 17Scaling Factor: 100 - Clients: 800 - Mode: Read Only - Average Latencyv6.11v6.12v6.13 7 Dec0.03760.07520.11280.15040.188SE +/- 0.001, N = 3SE +/- 0.000, N = 3SE +/- 0.000, N = 30.1670.1660.1651. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpq -lpgcommon -lpgport -lm

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL 17Scaling Factor: 100 - Clients: 800 - Mode: Read Onlyv6.11v6.12v6.13 7 Dec1000K2000K3000K4000K5000KSE +/- 11484.93, N = 3SE +/- 9902.78, N = 3SE +/- 8981.60, N = 34783648481562548292551. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpq -lpgcommon -lpgport -lm

OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL 17Scaling Factor: 100 - Clients: 1000 - Mode: Read Only - Average Latencyv6.11v6.13 7 Decv6.120.04880.09760.14640.19520.244SE +/- 0.001, N = 3SE +/- 0.001, N = 3SE +/- 0.001, N = 30.2170.2110.2111. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpq -lpgcommon -lpgport -lm

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL 17Scaling Factor: 100 - Clients: 1000 - Mode: Read Onlyv6.11v6.12v6.13 7 Dec1000K2000K3000K4000K5000KSE +/- 23812.59, N = 3SE +/- 18553.35, N = 3SE +/- 9098.55, N = 34621379473275347355781. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpq -lpgcommon -lpgport -lm

Palabos

OpenBenchmarking.orgMega Site Updates Per Second, More Is BetterPalabos 2.3Grid Size: 500v6.11v6.13 7 Decv6.122004006008001000SE +/- 40.67, N = 12SE +/- 5.28, N = 3SE +/- 1.19, N = 3529.23764.02775.111. (CXX) g++ options: -std=c++17 -pedantic -O3 -rdynamic -lcrypto -lcurl -lsz -lz -ldl -lm

Apache Cassandra

OpenBenchmarking.orgOp/s, More Is BetterApache Cassandra 5.0Test: Writesv6.11v6.13 7 Decv6.12110K220K330K440K550KSE +/- 5140.53, N = 3SE +/- 4648.88, N = 3SE +/- 1282.82, N = 3483952487743491186

ClickHouse

ClickHouse is an open-source, high performance OLAP data management system. This test profile uses ClickHouse's standard benchmark recommendations per https://clickhouse.com/docs/en/operations/performance-test/ / https://github.com/ClickHouse/ClickBench/tree/main/clickhouse with the 100 million rows web analytics dataset. The reported value is the query processing time using the geometric mean of all separate queries performed as an aggregate. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgQueries Per Minute, Geo Mean, More Is BetterClickHouse 22.12.3.5100M Rows Hits Dataset, Third Runv6.12v6.11v6.13 7 Dec2004006008001000SE +/- 7.60, N = 3SE +/- 4.79, N = 3SE +/- 2.96, N = 3799.99801.19807.27MIN: 64.94 / MAX: 10000MIN: 65.5 / MAX: 8571.43MIN: 66.01 / MAX: 8571.43

OpenBenchmarking.orgQueries Per Minute, Geo Mean, More Is BetterClickHouse 22.12.3.5100M Rows Hits Dataset, Second Runv6.12v6.13 7 Decv6.112004006008001000SE +/- 7.42, N = 3SE +/- 8.62, N = 3SE +/- 4.73, N = 3795.68803.49804.06MIN: 65.29 / MAX: 7500MIN: 66.08 / MAX: 10000MIN: 65.15 / MAX: 10000

OpenBenchmarking.orgQueries Per Minute, Geo Mean, More Is BetterClickHouse 22.12.3.5100M Rows Hits Dataset, First Run / Cold Cachev6.12v6.11v6.13 7 Dec2004006008001000SE +/- 5.98, N = 3SE +/- 2.38, N = 3SE +/- 2.84, N = 3765.59770.47774.01MIN: 63.29 / MAX: 7500MIN: 65.15 / MAX: 8571.43MIN: 65.86 / MAX: 7500

Timed LLVM Compilation

This test times how long it takes to compile/build the LLVM compiler stack. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed LLVM Compilation 16.0Build System: Ninjav6.11v6.13 7 Decv6.1220406080100SE +/- 0.11, N = 3SE +/- 0.19, N = 3SE +/- 0.06, N = 3101.82101.73101.60

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: Tradebeansv6.11v6.12v6.13 7 Dec11002200330044005500SE +/- 48.79, N = 15SE +/- 76.59, N = 15SE +/- 100.21, N = 12499949684957

Llamafile

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.8.16Model: wizardcoder-python-34b-v1.0.Q6_K - Test: Prompt Processing 2048v6.11v6.12v6.13 7 Dec3K6K9K12K15KSE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3122881228812288

Stress-NG

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.17.08Test: Semaphoresv6.11v6.13 7 Decv6.1250M100M150M200M250MSE +/- 2150155.18, N = 15SE +/- 1084579.30, N = 3SE +/- 3074547.79, N = 3233159294.14246944948.46251549880.011. (CXX) g++ options: -O2 -std=gnu99 -lc -lm

Memcached

Memcached is a high performance, distributed memory object caching system. This Memcached test profiles makes use of memtier_benchmark for excuting this CPU/memory-focused server benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOps/sec, More Is BetterMemcached 1.6.19Set To Get Ratio: 1:100v6.11v6.12v6.13 7 Dec3M6M9M12M15MSE +/- 13756.93, N = 3SE +/- 15384.68, N = 3SE +/- 24482.38, N = 312696519.8513294800.5913514659.671. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre

OpenBenchmarking.orgOps/sec, More Is BetterMemcached 1.6.19Set To Get Ratio: 1:10v6.12v6.13 7 Decv6.111.6M3.2M4.8M6.4M8MSE +/- 30399.87, N = 3SE +/- 37597.18, N = 3SE +/- 35717.40, N = 37297642.217400996.987421710.241. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre

OpenBenchmarking.orgOps/sec, More Is BetterMemcached 1.6.19Set To Get Ratio: 1:5v6.12v6.13 7 Decv6.11900K1800K2700K3600K4500KSE +/- 16085.40, N = 3SE +/- 28528.57, N = 3SE +/- 26467.49, N = 34064048.144099640.314115352.491. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre

Stress-NG

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.17.08Test: Context Switchingv6.11v6.13 7 Decv6.1212M24M36M48M60MSE +/- 247154.69, N = 13SE +/- 159927.59, N = 3SE +/- 64836.17, N = 336748330.5453038526.0055755486.321. (CXX) g++ options: -O2 -std=gnu99 -lc -lm

Laghos

OpenBenchmarking.orgMajor Kernels Total Rate, More Is BetterLaghos 3.1Test: Sedov Blast Wave, ube_922_hex.meshv6.11v6.12v6.13 7 Dec120240360480600SE +/- 3.27, N = 3SE +/- 2.47, N = 3SE +/- 2.95, N = 3561.91565.10565.141. (CXX) g++ options: -O3 -std=c++11 -lmfem -lHYPRE -lmetis -lrt -lmpi_cxx -lmpi

Llamafile

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.8.16Model: mistral-7b-instruct-v0.2.Q5_K_M - Test: Prompt Processing 2048v6.11v6.12v6.13 7 Dec7K14K21K28K35KSE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3327683276832768

Blender

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.3Blend File: Pabellon Barcelona - Compute: CPU-Onlyv6.11v6.12v6.13 7 Dec1122334455SE +/- 0.05, N = 3SE +/- 0.08, N = 3SE +/- 0.02, N = 347.9047.7047.39

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.3Blend File: Classroom - Compute: CPU-Onlyv6.11v6.12v6.13 7 Dec1020304050SE +/- 0.01, N = 3SE +/- 0.05, N = 3SE +/- 0.04, N = 342.0341.7441.49

SVT-AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.3Encoder Mode: Preset 5 - Input: Bosphorus 4Kv6.11v6.12v6.13 7 Dec1428425670SE +/- 0.07, N = 3SE +/- 0.12, N = 3SE +/- 0.08, N = 360.1760.3860.821. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

Llamafile

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.8.16Model: wizardcoder-python-34b-v1.0.Q6_K - Test: Text Generation 128v6.13 7 Decv6.12v6.1148121620SE +/- 0.02, N = 3SE +/- 0.04, N = 3SE +/- 0.02, N = 315.7715.8215.90

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.8.16Model: wizardcoder-python-34b-v1.0.Q6_K - Test: Prompt Processing 1024v6.11v6.12v6.13 7 Dec13002600390052006500SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3614461446144

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: Tradesoapv6.11v6.13 7 Decv6.127001400210028003500SE +/- 24.92, N = 3SE +/- 31.71, N = 5SE +/- 21.97, N = 15307730493035

Laghos

OpenBenchmarking.orgMajor Kernels Total Rate, More Is BetterLaghos 3.1Test: Triple Point Problemv6.11v6.13 7 Decv6.1270140210280350SE +/- 3.08, N = 3SE +/- 3.23, N = 3SE +/- 1.53, N = 3297.99300.72303.111. (CXX) g++ options: -O3 -std=c++11 -lmfem -lHYPRE -lmetis -lrt -lmpi_cxx -lmpi

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: Apache Lucene Search Enginev6.12v6.11v6.13 7 Dec9001800270036004500SE +/- 47.07, N = 3SE +/- 43.31, N = 15SE +/- 42.57, N = 3432641203631

OpenVINO GenAI

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO GenAI 2024.5Model: Falcon-7b-instruct-int4-ov - Device: CPU - Time Per Output Tokenv6.12v6.13 7 Decv6.1148121620SE +/- 0.17, N = 6SE +/- 0.17, N = 3SE +/- 0.19, N = 317.8017.7717.63

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO GenAI 2024.5Model: Falcon-7b-instruct-int4-ov - Device: CPU - Time To First Tokenv6.11v6.13 7 Decv6.12714212835SE +/- 0.27, N = 3SE +/- 0.44, N = 3SE +/- 0.26, N = 630.3930.1830.00

OpenBenchmarking.orgtokens/s, More Is BetterOpenVINO GenAI 2024.5Model: Falcon-7b-instruct-int4-ov - Device: CPUv6.12v6.13 7 Decv6.111326395265SE +/- 0.55, N = 6SE +/- 0.54, N = 3SE +/- 0.63, N = 356.2156.2856.73

Stress-NG

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.17.08Test: NUMAv6.11v6.12v6.13 7 Dec5001000150020002500SE +/- 6.78, N = 3SE +/- 1.02, N = 3SE +/- 10.25, N = 32071.982298.652385.591. (CXX) g++ options: -O2 -std=gnu99 -lc -lm

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.17.08Test: Mixed Schedulerv6.11v6.13 7 Decv6.1220K40K60K80K100KSE +/- 161.54, N = 3SE +/- 282.83, N = 3SE +/- 215.02, N = 379773.3781582.5882674.981. (CXX) g++ options: -O2 -std=gnu99 -lc -lm

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.17.08Test: Futexv6.11v6.12v6.13 7 Dec900K1800K2700K3600K4500KSE +/- 17548.84, N = 3SE +/- 15124.18, N = 3SE +/- 31680.82, N = 34154651.854313954.944398887.771. (CXX) g++ options: -O2 -std=gnu99 -lc -lm

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.17.08Test: MEMFDv6.12v6.11v6.13 7 Dec5001000150020002500SE +/- 2.72, N = 3SE +/- 4.79, N = 3SE +/- 3.27, N = 32010.992054.702125.131. (CXX) g++ options: -O2 -std=gnu99 -lc -lm

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.17.08Test: Socket Activityv6.13 7 Decv6.11v6.1210K20K30K40K50KSE +/- 42.06, N = 3SE +/- 29.82, N = 3SE +/- 68.22, N = 347238.7047424.8747501.041. (CXX) g++ options: -O2 -std=gnu99 -lc -lm

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.17.08Test: Mutexv6.13 7 Decv6.11v6.1210M20M30M40M50MSE +/- 188378.96, N = 3SE +/- 181454.22, N = 3SE +/- 442253.04, N = 343186401.8445376718.9445398166.141. (CXX) g++ options: -O2 -std=gnu99 -lc -lm

Llamafile

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.8.16Model: mistral-7b-instruct-v0.2.Q5_K_M - Test: Text Generation 128v6.13 7 Decv6.12v6.111530456075SE +/- 0.36, N = 3SE +/- 0.47, N = 3SE +/- 0.37, N = 365.4567.7168.13

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.8.16Model: TinyLlama-1.1B-Chat-v1.0.BF16 - Test: Text Generation 128v6.13 7 Decv6.12v6.114080120160200SE +/- 1.45, N = 15SE +/- 1.26, N = 3SE +/- 1.11, N = 3156.82159.60159.90

Timed Linux Kernel Compilation

This test times how long it takes to build the Linux kernel in a default configuration (defconfig) for the architecture being tested or alternatively an allmodconfig for building all possible kernel modules for the build. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Linux Kernel Compilation 6.8Build: defconfigv6.13 7 Decv6.12v6.11510152025SE +/- 0.26, N = 4SE +/- 0.31, N = 3SE +/- 0.25, N = 421.8621.8521.79

OpenVINO GenAI

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO GenAI 2024.5Model: Gemma-7b-int4-ov - Device: CPU - Time Per Output Tokenv6.13 7 Decv6.12v6.11510152025SE +/- 0.11, N = 3SE +/- 0.04, N = 3SE +/- 0.06, N = 320.5420.3720.36

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO GenAI 2024.5Model: Gemma-7b-int4-ov - Device: CPU - Time To First Tokenv6.13 7 Decv6.11v6.12714212835SE +/- 0.18, N = 3SE +/- 0.08, N = 3SE +/- 0.04, N = 330.4830.3930.27

OpenBenchmarking.orgtokens/s, More Is BetterOpenVINO GenAI 2024.5Model: Gemma-7b-int4-ov - Device: CPUv6.13 7 Decv6.12v6.111122334455SE +/- 0.27, N = 3SE +/- 0.10, N = 3SE +/- 0.15, N = 348.7049.0949.13

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: Avrora AVR Simulation Frameworkv6.11v6.12v6.13 7 Dec5001000150020002500SE +/- 16.90, N = 3SE +/- 15.56, N = 13SE +/- 20.72, N = 15235423032283

Primesieve

OpenBenchmarking.orgSeconds, Fewer Is BetterPrimesieve 12.6Length: 1e13v6.11v6.12v6.13 7 Dec612182430SE +/- 0.04, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 325.0424.9724.941. (CXX) g++ options: -O3

Llamafile

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.8.16Model: mistral-7b-instruct-v0.2.Q5_K_M - Test: Prompt Processing 1024v6.11v6.12v6.13 7 Dec4K8K12K16K20KSE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3163841638416384

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.8.16Model: wizardcoder-python-34b-v1.0.Q6_K - Test: Text Generation 16v6.13 7 Decv6.11v6.1248121620SE +/- 0.73, N = 12SE +/- 0.74, N = 12SE +/- 0.74, N = 1214.9014.9815.01

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: Eclipsev6.11v6.13 7 Decv6.1214002800420056007000SE +/- 14.75, N = 3SE +/- 23.78, N = 3SE +/- 33.39, N = 3634963316280

Llamafile

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.8.16Model: Llama-3.2-3B-Instruct.Q6_K - Test: Prompt Processing 2048v6.11v6.12v6.13 7 Dec7K14K21K28K35KSE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3327683276832768

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: H2 Database Enginev6.11v6.12v6.13 7 Dec5001000150020002500SE +/- 28.09, N = 3SE +/- 13.65, N = 3SE +/- 18.21, N = 3210420492010

Blender

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.3Blend File: Fishy Cat - Compute: CPU-Onlyv6.11v6.12v6.13 7 Dec510152025SE +/- 0.12, N = 3SE +/- 0.06, N = 3SE +/- 0.06, N = 321.6621.4921.35

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: jMonkeyEnginev6.12v6.13 7 Decv6.1115003000450060007500SE +/- 4.84, N = 3SE +/- 1.86, N = 3SE +/- 1.86, N = 3679767936791

Blender

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.3Blend File: Junkshop - Compute: CPU-Onlyv6.11v6.12v6.13 7 Dec510152025SE +/- 0.04, N = 3SE +/- 0.06, N = 3SE +/- 0.06, N = 320.2420.2120.20

Llamafile

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.8.16Model: wizardcoder-python-34b-v1.0.Q6_K - Test: Prompt Processing 512v6.11v6.12v6.13 7 Dec7001400210028003500SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3307230723072

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: PMD Source Code Analyzerv6.12v6.13 7 Decv6.1130060090012001500SE +/- 23.97, N = 15SE +/- 18.53, N = 15SE +/- 5.90, N = 3116710891085

Llamafile

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.8.16Model: Llama-3.2-3B-Instruct.Q6_K - Test: Text Generation 128v6.13 7 Decv6.12v6.11306090120150SE +/- 1.53, N = 3SE +/- 1.04, N = 3SE +/- 0.87, N = 3109.60111.98112.56

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: Apache Kafkav6.11v6.13 7 Decv6.1211002200330044005500SE +/- 7.33, N = 3SE +/- 3.53, N = 3SE +/- 26.64, N = 3507050575045

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: Apache Lucene Search Indexv6.11v6.13 7 Decv6.125001000150020002500SE +/- 16.64, N = 3SE +/- 25.32, N = 3SE +/- 8.39, N = 3229422782258

Blender

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.3Blend File: BMW27 - Compute: CPU-Onlyv6.11v6.12v6.13 7 Dec48121620SE +/- 0.00, N = 3SE +/- 0.05, N = 3SE +/- 0.04, N = 315.1915.0815.02

Llamafile

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.8.16Model: mistral-7b-instruct-v0.2.Q5_K_M - Test: Text Generation 16v6.13 7 Decv6.12v6.111428425670SE +/- 1.70, N = 12SE +/- 1.75, N = 12SE +/- 1.73, N = 1264.1564.8064.87

OpenVINO GenAI

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO GenAI 2024.5Model: TinyLlama-1.1B-Chat-v1.0 - Device: CPU - Time Per Output Tokenv6.13 7 Decv6.12v6.113691215SE +/- 0.04, N = 3SE +/- 0.04, N = 3SE +/- 0.09, N = 312.8812.7912.76

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO GenAI 2024.5Model: TinyLlama-1.1B-Chat-v1.0 - Device: CPU - Time To First Tokenv6.13 7 Decv6.12v6.1148121620SE +/- 0.08, N = 3SE +/- 0.05, N = 3SE +/- 0.19, N = 316.0515.8915.86

OpenBenchmarking.orgtokens/s, More Is BetterOpenVINO GenAI 2024.5Model: TinyLlama-1.1B-Chat-v1.0 - Device: CPUv6.13 7 Decv6.12v6.1120406080100SE +/- 0.28, N = 3SE +/- 0.23, N = 3SE +/- 0.57, N = 377.6778.1878.36

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: Jythonv6.11v6.13 7 Decv6.128001600240032004000SE +/- 6.67, N = 3SE +/- 11.00, N = 3SE +/- 5.33, N = 3378237633755

SVT-AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.3Encoder Mode: Preset 8 - Input: Bosphorus 4Kv6.11v6.12v6.13 7 Dec4080120160200SE +/- 1.84, N = 3SE +/- 1.77, N = 3SE +/- 0.64, N = 3199.21200.32200.821. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

Llamafile

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.8.16Model: TinyLlama-1.1B-Chat-v1.0.BF16 - Test: Prompt Processing 2048v6.11v6.12v6.13 7 Dec7K14K21K28K35KSE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3327683276832768

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: BioJava Biological Data Frameworkv6.12v6.11v6.13 7 Dec9001800270036004500SE +/- 18.98, N = 3SE +/- 11.89, N = 3SE +/- 22.00, N = 3407140584019

Llamafile

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.8.16Model: mistral-7b-instruct-v0.2.Q5_K_M - Test: Prompt Processing 512v6.11v6.12v6.13 7 Dec2K4K6K8K10KSE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3819281928192

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: Spring Bootv6.11v6.13 7 Decv6.126001200180024003000SE +/- 17.89, N = 3SE +/- 26.77, N = 3SE +/- 20.60, N = 3272426512259

OpenVINO GenAI

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO GenAI 2024.5Model: Phi-3-mini-128k-instruct-int4-ov - Device: CPU - Time Per Output Tokenv6.12v6.11v6.13 7 Dec48121620SE +/- 0.02, N = 3SE +/- 0.10, N = 3SE +/- 0.18, N = 315.1915.1815.17

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO GenAI 2024.5Model: Phi-3-mini-128k-instruct-int4-ov - Device: CPU - Time To First Tokenv6.12v6.11v6.13 7 Dec510152025SE +/- 0.37, N = 3SE +/- 0.24, N = 3SE +/- 0.43, N = 322.7322.6622.38

OpenBenchmarking.orgtokens/s, More Is BetterOpenVINO GenAI 2024.5Model: Phi-3-mini-128k-instruct-int4-ov - Device: CPUv6.12v6.11v6.13 7 Dec1530456075SE +/- 0.10, N = 3SE +/- 0.41, N = 3SE +/- 0.79, N = 365.8265.8965.91

Llamafile

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.8.16Model: Llama-3.2-3B-Instruct.Q6_K - Test: Prompt Processing 1024v6.11v6.12v6.13 7 Dec4K8K12K16K20KSE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3163841638416384

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.8.16Model: Llama-3.2-3B-Instruct.Q6_K - Test: Text Generation 16v6.13 7 Decv6.12v6.1120406080100SE +/- 2.52, N = 14SE +/- 2.55, N = 12SE +/- 2.48, N = 12102.21107.95108.22

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: GraphChiv6.13 7 Decv6.11v6.125001000150020002500SE +/- 6.00, N = 3SE +/- 13.86, N = 3SE +/- 5.51, N = 3210220932083

Llamafile

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.8.16Model: wizardcoder-python-34b-v1.0.Q6_K - Test: Prompt Processing 256v6.11v6.12v6.13 7 Dec30060090012001500SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3153615361536

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.8.16Model: TinyLlama-1.1B-Chat-v1.0.BF16 - Test: Text Generation 16v6.13 7 Decv6.11v6.12306090120150SE +/- 4.35, N = 12SE +/- 4.56, N = 12SE +/- 4.63, N = 12147.98152.07152.89

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.8.16Model: mistral-7b-instruct-v0.2.Q5_K_M - Test: Prompt Processing 256v6.11v6.12v6.13 7 Dec9001800270036004500SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3409640964096

SVT-AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.3Encoder Mode: Preset 13 - Input: Bosphorus 4Kv6.11v6.12v6.13 7 Dec100200300400500SE +/- 3.40, N = 3SE +/- 1.95, N = 3SE +/- 0.44, N = 3466.61468.67480.911. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

Llamafile

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.8.16Model: TinyLlama-1.1B-Chat-v1.0.BF16 - Test: Prompt Processing 1024v6.11v6.12v6.13 7 Dec4K8K12K16K20KSE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3163841638416384

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: Apache Tomcatv6.13 7 Decv6.11v6.122004006008001000SE +/- 2.00, N = 3SE +/- 7.81, N = 3SE +/- 3.51, N = 3105510531038

Llamafile

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.8.16Model: Llama-3.2-3B-Instruct.Q6_K - Test: Prompt Processing 512v6.11v6.12v6.13 7 Dec2K4K6K8K10KSE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3819281928192

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: Zxing 1D/2D Barcode Image Processingv6.12v6.11v6.13 7 Dec120240360480600SE +/- 4.38, N = 9SE +/- 4.26, N = 3SE +/- 5.55, N = 3551522493

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: Batik SVG Toolkitv6.13 7 Decv6.12v6.112004006008001000SE +/- 7.55, N = 3SE +/- 4.00, N = 3SE +/- 3.61, N = 3938937923

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: Apache Xalan XSLTv6.11v6.13 7 Decv6.12160320480640800SE +/- 6.06, N = 3SE +/- 8.54, N = 3SE +/- 1.76, N = 3743742741

Llamafile

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.8.16Model: Llama-3.2-3B-Instruct.Q6_K - Test: Prompt Processing 256v6.11v6.12v6.13 7 Dec9001800270036004500SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3409640964096

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.8.16Model: TinyLlama-1.1B-Chat-v1.0.BF16 - Test: Prompt Processing 512v6.11v6.12v6.13 7 Dec2K4K6K8K10KSE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3819281928192

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: FOP Print Formatterv6.11v6.13 7 Decv6.1270140210280350SE +/- 0.67, N = 3SE +/- 1.20, N = 3SE +/- 2.08, N = 3339338335

Primesieve

OpenBenchmarking.orgSeconds, Fewer Is BetterPrimesieve 12.6Length: 1e12v6.11v6.12v6.13 7 Dec0.44980.89961.34941.79922.249SE +/- 0.003, N = 3SE +/- 0.001, N = 3SE +/- 0.002, N = 31.9991.9901.9881. (CXX) g++ options: -O3

Llamafile

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.8.16Model: TinyLlama-1.1B-Chat-v1.0.BF16 - Test: Prompt Processing 256v6.11v6.12v6.13 7 Dec9001800270036004500SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3409640964096

110 Results Shown

Apache CouchDB:
  300 - 3000 - 30
  500 - 1000 - 30
RELION
Palabos
MariaDB:
  oltp_update_index - 128
  oltp_read_only - 128
  oltp_read_write - 128
  oltp_write_only - 128
Timed Linux Kernel Compilation
OpenSSL:
  ChaCha20
  ChaCha20-Poly1305
  AES-256-GCM
  AES-128-GCM
Timed LLVM Compilation
Blender
Apache CouchDB
SVT-AV1
PostgreSQL:
  100 - 1000 - Read Write - Average Latency
  100 - 1000 - Read Write
  100 - 800 - Read Write - Average Latency
  100 - 800 - Read Write
  100 - 800 - Read Only - Average Latency
  100 - 800 - Read Only
  100 - 1000 - Read Only - Average Latency
  100 - 1000 - Read Only
Palabos
Apache Cassandra
ClickHouse:
  100M Rows Hits Dataset, Third Run
  100M Rows Hits Dataset, Second Run
  100M Rows Hits Dataset, First Run / Cold Cache
Timed LLVM Compilation
DaCapo Benchmark
Llamafile
Stress-NG
Memcached:
  1:100
  1:10
  1:5
Stress-NG
Laghos
Llamafile
Blender:
  Pabellon Barcelona - CPU-Only
  Classroom - CPU-Only
SVT-AV1
Llamafile:
  wizardcoder-python-34b-v1.0.Q6_K - Text Generation 128
  wizardcoder-python-34b-v1.0.Q6_K - Prompt Processing 1024
DaCapo Benchmark
Laghos
DaCapo Benchmark
OpenVINO GenAI:
  Falcon-7b-instruct-int4-ov - CPU - Time Per Output Token
  Falcon-7b-instruct-int4-ov - CPU - Time To First Token
  Falcon-7b-instruct-int4-ov - CPU
Stress-NG:
  NUMA
  Mixed Scheduler
  Futex
  MEMFD
  Socket Activity
  Mutex
Llamafile:
  mistral-7b-instruct-v0.2.Q5_K_M - Text Generation 128
  TinyLlama-1.1B-Chat-v1.0.BF16 - Text Generation 128
Timed Linux Kernel Compilation
OpenVINO GenAI:
  Gemma-7b-int4-ov - CPU - Time Per Output Token
  Gemma-7b-int4-ov - CPU - Time To First Token
  Gemma-7b-int4-ov - CPU
DaCapo Benchmark
Primesieve
Llamafile:
  mistral-7b-instruct-v0.2.Q5_K_M - Prompt Processing 1024
  wizardcoder-python-34b-v1.0.Q6_K - Text Generation 16
DaCapo Benchmark
Llamafile
DaCapo Benchmark
Blender
DaCapo Benchmark
Blender
Llamafile
DaCapo Benchmark
Llamafile
DaCapo Benchmark:
  Apache Kafka
  Apache Lucene Search Index
Blender
Llamafile
OpenVINO GenAI:
  TinyLlama-1.1B-Chat-v1.0 - CPU - Time Per Output Token
  TinyLlama-1.1B-Chat-v1.0 - CPU - Time To First Token
  TinyLlama-1.1B-Chat-v1.0 - CPU
DaCapo Benchmark
SVT-AV1
Llamafile
DaCapo Benchmark
Llamafile
DaCapo Benchmark
OpenVINO GenAI:
  Phi-3-mini-128k-instruct-int4-ov - CPU - Time Per Output Token
  Phi-3-mini-128k-instruct-int4-ov - CPU - Time To First Token
  Phi-3-mini-128k-instruct-int4-ov - CPU
Llamafile:
  Llama-3.2-3B-Instruct.Q6_K - Prompt Processing 1024
  Llama-3.2-3B-Instruct.Q6_K - Text Generation 16
DaCapo Benchmark
Llamafile:
  wizardcoder-python-34b-v1.0.Q6_K - Prompt Processing 256
  TinyLlama-1.1B-Chat-v1.0.BF16 - Text Generation 16
  mistral-7b-instruct-v0.2.Q5_K_M - Prompt Processing 256
SVT-AV1
Llamafile
DaCapo Benchmark
Llamafile
DaCapo Benchmark:
  Zxing 1D/2D Barcode Image Processing
  Batik SVG Toolkit
  Apache Xalan XSLT
Llamafile:
  Llama-3.2-3B-Instruct.Q6_K - Prompt Processing 256
  TinyLlama-1.1B-Chat-v1.0.BF16 - Prompt Processing 512
DaCapo Benchmark
Primesieve
Llamafile